基于术前PIV、NLR和CEA水平对胃癌根治术后预后列线图预测模型的构建及评价  

Construction and evaluation of nomogram prediction model for prognosis after radical gastrectomy based on preoperative PIV,NLR and CEA level

作　　者：陆通 房钰 刘浩楠 屈鹏飞 韩正祥 吴克俭 LU Tong;FANG Yu;LIU Haonan;QU Pengfei;HAN Zhengxiang;WU Kejian(The First Clinical Medical College of Xuzhou Medical University,Jiangsu Xuzhou 221002,China;Department of Gastroenterology,the Second Affiliated Hospital of Xuzhou Medical University,Jiangsu Xuzhou 221002,China;Department of Oncology,Affiliated Hospital of Xuzhou Medical University,Jiangsu Xuzhou 221002,China;Department of Gastroenterology,Affiliated Hospital of Xuzhou Medical University,Jiangsu Xuzhou 221002,China)

机构地区：[1]徐州医科大学第一临床医学院,江苏徐州221002 [2]徐州医科大学第二附属医院消化内科,江苏徐州221002 [3]徐州医科大学附属医院肿瘤内科,江苏徐州221002 [4]徐州医科大学附属医院消化内科,江苏徐州221002

出　　处：《现代肿瘤医学》2023年第10期1869-1875,共7页Journal of Modern Oncology

基　　金：2020年江苏省高层次卫生人才“六个一工程”拔尖人才科研项目(编号:LGY2020006)。

摘　　要：目的:基于术前泛免疫炎症(PIV)、中性粒细胞与淋巴细胞比值(NLR)以及癌胚抗原(CEA)水平探讨胃癌根治术后预后的影响因素并建立列线图预后预测模型。方法:回顾性分析2016年03月至2019年11月在徐州医科大学附属医院普外科行胃癌根治术的384例胃癌患者的临床病理资料,采用受试者工作特征(ROC)曲线分析术前PIV、NLR、CEA水平预测总生存期(OS)的最佳截断值,并根据PIV的最佳截断值进行分组。采用χ^(2)检验分析不同PIV水平与患者临床病理特征的关系。使用Kaplan-Meier法和Log-rank检验分析不同临床病理特征对患者OS的影响,多因素Cox回归分析患者预后的独立影响因素。使用R4.1.1软件绘制胃癌根治术后患者1、3、5年OS的列线图预测模型,并评价预测模型的效能,然后使用X-tile软件根据列线图风险得分将该模型分层进一步探讨该模型的临床应用价值。结果:ROC曲线分析结果显示,PIV、NLR、CEA曲线下面积(AUC)分别为0.627、0.584、0.590,最佳截断值分别为236.8、1.98、4.93 ng/mL。PIV与年龄、肿瘤最大直径、肿瘤浸润深度、淋巴结转移、TNM分期、神经或脉管侵犯、术前NLR水平相关(P<0.05)。多因素Cox回归分析显示,年龄、肿瘤浸润深度、神经或脉管侵犯、PIV、NLR、CEA为胃癌根治术后患者1、3、5年OS的独立影响因素(P<0.05)。构建包含以上独立危险因素的列线图预测模型,模型内部验证一致性指数(C指数)分别为0.797、0.805、0.780,校正曲线提示该模型区分度良好,低风险患者的OS明显优于中、高风险组(P<0.001)。结论:PIV、NLR、CEA对于胃癌预后有较好的预测价值,基于PIV、NLR、CEA水平及胃癌相关病理资料构建的列线图模型对于临床有较高的指导意义。Objective:Based on preoperative level of pan immune inflammation value(PIV),neutrophil-to-lymphocyte ratio(NLR)and carcinoembryonic antigen(CEA),to explore the influencing factors of prognosis after radical gastrectomy for gastric cancer,and to establish a nomogram prediction model.Methods:The clinical and pathological data of 384 patients with gastric cancer who underwent radical gastrectomy in the department of general surgery of Affiliated Hospital of Xuzhou Medical University from March 2016 to November 2019 were retrospectively analyzed.The best cutoff values of preoperative PIV,NLR and CEA level for predicting overall survival(OS)were analyzed by receiver operator characteristic(ROC)curve,and patients were divided into groups according to the best cutoff values of PIV.χ^(2) test was used to analyze the relationship between different PIV levels and clinicopathological features of gastric cancer patients.Kaplan-Meier method and Log-rank test were used to analyze the influence of different clinicopathological features on patients'OS,and multivariate Cox regression was used to analyze the independent influencing factors of patients'prognosis.R4.1.1 software was used to draw the nomogram prediction model of OS in patients with gastric cancer after radical gastrectomy for 1,3 and 5 years,and the effectiveness of the prediction model was evaluated.Then X-tile software was used to stratify the model according to the nomogram risk score to further explore the clinical application value of the model.Results:ROC curve analysis showed that the areas under PIV,NLR and CEA curves(AUC)were 0.627,0.584 and 0.590,respectively,and the best cut-off values were 236.8,1.98 and 4.93 ng/mL,respectively.PIV was related to age,tumor maximum diameter,tumor invasion depth,lymph node metastasis,TNM stage,nerve or vascular invasion and preoperative NLR level(P<0.05).Multivariate Cox regression analysis showed that age,depth of tumor invasion,nerve or vascular invasion,PIV,NLR and CEA were independent influencing factors of 1,3,5 years

关 键 词：胃癌 PIV 列线图 预后模型 

分 类 号：R735.2[医药卫生—肿瘤]